Coercive Control in Relationships: Statistics, Warning Signs and Support

ONS estimated 3.8 million adults experienced domestic abuse in England and Wales in the year to March 2025. Learn what the data show and how control can develop.

What the latest comparable survey measures

The Office for National Statistics estimated that 3.8 million people aged 16 and over in England and Wales experienced domestic abuse in the year ending March 2025, equivalent to 7.8%. The source is the Crime Survey for England and Wales, not simply police reports. The distinction matters because many incidents never reach the police, and administrative totals reflect recording practices as well as behaviour. The survey estimate provides a broader measure of experience within its defined population.

ONS says a new set of survey questions was introduced from the year ending March 2024, so the newer estimates should not be directly compared with older figures as if methodology were unchanged. Responsible analysis must say this explicitly. Otherwise an apparent trend could be created by a new measurement instrument rather than a real change in prevalence. The report provides serious context, but it cannot determine whether a particular relationship is abusive without understanding the conduct involved.

Coercive control is a pattern, not one argument

Conflict and disappointment occur in many healthy relationships; coercive control involves a pattern of behaviour used to restrict another person’s freedom and independence. It can include isolating someone from friends, monitoring communications, threatening consequences for ordinary decisions or restricting access to money. The specific legal definition depends on jurisdiction, and a public-health description is not identical to a criminal-law assessment.

The difference is often evident in the effect on daily choice. Can someone visit family without fearing retaliation? Can they say no to sex? Can they keep their own phone, finances and friendships? Do disagreements end with mutual negotiation, or does one person enforce compliance? A single example may not settle the answer, but recurring intimidation and restricted autonomy warrant serious concern. The person affected should be heard rather than challenged to prove a perfect pattern.

The numbers on different types of abuse

ONS estimated that in the year ending March 2025, emotional abuse affected 5.0% of people aged 16 and over in England and Wales, economic abuse 2.7% and threats 2.6%. These categories may overlap, so adding the percentages together would not give an overall abuse rate. This is a fundamental statistical point that is easily missed in headlines. Different experiences may co-occur within a relationship, and measured categories are shaped by the survey questions.

The data also showed different prevalence by sex: about 9.1% of women and 6.5% of men experienced domestic abuse in the past year. These figures do not deny anyone’s experience. They show why support systems need to recognise both unequal patterns and the reality that any individual can be harmed. A man seeking help should not be dismissed, and a woman seeking help should not be assumed to fit one stereotyped story.

Economic abuse and daily dependence

Economic abuse may involve preventing someone from working, controlling access to income, taking on debt in their name or demanding explanations for ordinary purchases. A couple deciding together how to budget is different from one partner using money to restrict the other’s movement or ability to leave. It is important not to label every unequal income arrangement abusive without evidence. The relevant question is who retains meaningful choice and whether financial rules are mutually agreed.

For someone affected, practical safety may include identifying where essential documents are held, speaking confidentially to specialist services and understanding financial options. However, generic advice to confront the person or immediately move money can increase risk in some circumstances. A support service can help plan safely for the individual situation. Protecting autonomy matters more than meeting an internet checklist.

Technology-facilitated monitoring

Digital control can involve demanding passwords, monitoring locations, reading messages or using threats to obtain intimate material. Partners may voluntarily share information, but voluntary sharing is not the same as being unable to refuse. A request that triggers anger or punishment when declined should be assessed in its wider context. Changing a password without planning may be unsafe if someone is being monitored closely, so specialist advice may be appropriate.

Technology can also provide vital access to help. A person concerned about monitoring might prefer a trusted device or private conversation with a support worker, depending on their circumstances. There is no universal digital escape plan that is safe for everyone. Articles should avoid instructing people to take high-risk actions without considering their immediate environment and available professional support.

Sexual consent within an established relationship

A long relationship, marriage or previous sexual experience never creates permanent sexual consent. Pressure can include threats, persistent bargaining or punishment after refusal. Someone who feels they must participate to avoid conflict may not be choosing freely. Legal definitions vary, but ethically the underlying principle is clear: both partners retain authority over their own bodies and may decline or stop.

When someone describes coercion, the first response should not be to insist on joint relationship counselling. Couple therapy may be inappropriate where there is fear or abuse because it can create additional risks. Individual specialist support can help assess options without pressuring the person to confront the other partner. The objective is safety and informed choice, not preserving a relationship at any cost.

What police data can and cannot establish

ONS recorded 49,557 offences of controlling or coercive behaviour in England and Wales in the year ending March 2025. This is a police-recorded offence count, not an estimate of how many people experienced coercive control. The survey questions do not perfectly match the legal definition, so the survey cannot simply be used to calculate prevalence of the specific offence. Mixing these two measures in one graphic without explanation would mislead readers.

Police totals may also be influenced by reporting confidence, police practice and recording changes. This is one reason careful reporting cites both administrative data and surveys with their limitations. A rise or fall in recorded offences does not, on its own, demonstrate the same rise or fall in underlying harm. Better analysis asks what each dataset is built to measure.

Recognising isolation and intimidation

One of the more subtle developments can be a gradual narrowing of ordinary life. A person may stop seeing friends because every visit produces accusations, or avoid disagreeing because arguments bring threats. Other people may notice the change before the affected individual can name it. Support should remain non-judgemental: demanding that someone leave immediately may ignore practical dangers, finances, children, housing and emotional dependence.

A friend can listen, avoid blaming the person for staying and offer to help contact a specialist service safely. Do not share private information or confront the alleged perpetrator without considering the consequences. Respect the individual’s decisions while being clear that control and intimidation are not ordinary relationship compromises. Safety planning works best with people who understand domestic abuse risks.

Finding support and deciding next steps

Anyone in immediate danger should contact emergency services. Specialist helplines and local domestic abuse organisations can explain options around safety, housing, finances and legal support. The National Domestic Abuse Helpline is one UK resource, but access and legal pathways vary across nations. A person may seek information without being ready to leave or report. That choice should be respected unless emergency intervention is necessary.

Readers should not use one article to diagnose a partner or decide a criminal case. They can use the evidence to recognise patterns, take concerns seriously and seek qualified support. The difference between healthy disagreement and control is not that one couple never argues; it is whether both people remain free to make choices without fear or punishment.

Why not every disagreement is coercive control

A couple can argue over money, social plans or privacy without one person attempting to dominate the other. The critical question is whether a pattern of intimidation, restrictions or consequences has emerged that removes meaningful choice. Someone who requests more time together is different from someone who threatens punishment for seeing friends. Distinguishing those situations protects the accuracy of the term coercive control. It also helps readers recognise why seemingly minor demands may be alarming when they form part of a sustained pattern. Individual safety assessments need context rather than a score from a quiz.

Economic abuse without obvious theft

Financial control can exist without a dramatic incident in which money is stolen. Someone may be discouraged from working, forced to account for every purchase or left unable to access funds needed to travel. In a healthy agreement, financial roles can be uneven but remain negotiable and transparent. The warning sign is not simply unequal earnings: it is control used to diminish independence or create fear. Advice to immediately challenge a controlling partner may be unsafe. A specialist support service can help consider documents, housing, debts and practical options according to the person’s circumstances.

Why multiple data series should not be combined

The Crime Survey for England and Wales measures experiences reported by sampled individuals, while police data count offences brought to police attention and recorded. The groups overlap, but neither is a simple subset of the other. Survey percentages for emotional abuse, economic abuse and threats also overlap within respondents. Adding them produces an inflated total. ONS further warns that its questions changed from March 2024, limiting comparison with older figures. A careful article should maintain these distinctions rather than using several statistics to create the most dramatic possible headline.

A safer response when devices may be monitored

A person who suspects monitoring may be at risk if they suddenly change passwords, disable location sharing or confront the other person. Those actions can be appropriate in some situations but require consideration of immediate safety. A trusted device, private support contact or specialist service may offer safer ways to explore options. Do not assume that an incognito browser hides all activity from someone with access to a device or network. Support should be individualised, not presented as a universal technical escape plan. Protecting the person takes priority over the elegance of a digital checklist.

Friends and family can help without taking control

Well-meaning friends sometimes demand an immediate breakup and become frustrated if the person stays. That reaction can reproduce the feeling of losing agency. Listen to what the person says, express concern about specific conduct and ask how you can support them safely. Avoid contacting the alleged abuser or disclosing private plans. Offer practical options such as accompanying someone to a support service if they want that. A person may need several conversations before making a decision. Their pace does not mean the harm is acceptable or that the concern was mistaken.

What to do when someone is in immediate danger

Where there is an immediate threat, emergency services are appropriate. Otherwise, specialist domestic abuse organisations can advise about safety, accommodation, financial concerns and legal options. Different UK nations have different services and rules, so a London-focused article should not pretend every local procedure applies across the country. Someone can seek confidential information without committing to a police report or immediate separation. The appropriate plan depends on risk, children, housing, finances and other circumstances. The goal is not to satisfy an outside observer’s deadline but to increase safety and choice.

Can someone experience economic abuse in a shared account?

Yes, but account structure alone does not demonstrate abuse. Many couples voluntarily use shared finances. Concern arises when a person is denied reasonable access, threatened for spending necessities or prevented from working or leaving. Consider whether decisions are mutually agreed and whether one partner can challenge them without fear. ONS measures economic abuse as a form of domestic abuse, but a survey category does not replace a personalised safety assessment. People experiencing control may benefit from confidential financial and domestic-abuse advice.

What if the person says they are not ready to leave?

Avoid assuming that a simple instruction to leave accounts for housing, children, immigration status, finances, threats or emotional attachment. A person may face increased danger around separation. Specialist services can help develop a safer plan. Friends can remain available and avoid shame or ultimatums. Acknowledging the risks does not normalise control; it respects that the individual must make decisions within real constraints. The appropriate role for an article is to identify support, not to prescribe an immediate confrontation.

Does monitoring a partner ever count as consent?

Two people may voluntarily choose to share locations or devices. The ethical difference lies in freedom to refuse or withdraw without punishment. Demanding access, using threats or covertly monitoring someone forms a different pattern. If a person suspects coercive monitoring, suddenly disabling it may have safety consequences. Seek specialist advice about safe communication and device use. No generic digital-security instructions should be treated as universally safe for people facing abuse. Context and risk matter more than a technically elegant workaround.

What about men who experience abuse?

ONS estimated that 6.5% of men in England and Wales experienced domestic abuse in the year ending March 2025, compared with 9.1% of women. Both figures reflect the survey’s definitions. The difference matters for public policy but does not make any individual less deserving of help. Support services should be accessible to victims across sexes, ages and backgrounds. Readers should avoid stereotypes that men cannot be intimidated or that women are always safe from coercion. The first response to disclosure should be to take it seriously.

How is coercive control different from poor communication?

A couple can communicate badly and still retain equality of choice. Coercive control involves conduct that systematically restricts another person’s autonomy or creates fear of consequences. Someone who apologises after an argument but continues to isolate their partner or control finances may be demonstrating a pattern, not merely a lack of listening skills. Conversely, one sharp disagreement does not by itself establish a criminal offence. The person concerned can seek specialist support to examine the broader context rather than relying on an online quiz that assigns a label automatically.

Are recorded crimes a direct measure of prevalence?

No. Police-recorded offences represent incidents that entered and were recorded by the criminal justice system. Many experiences of abuse are not reported, while recording policy can change. The ONS survey is designed to estimate prevalence in a sampled population. Its figures and police counts answer related but different questions. The 49,557 recorded offences of controlling or coercive behaviour in the year ending March 2025 should not be rewritten as 49,557 victims or as the total extent of coercion. Accurate definitions matter especially when the subject is serious.

What should a publication do for someone seeking help?

Offer clear descriptions of concerning behaviour, identify specialist support and distinguish urgent danger from general relationship dissatisfaction. Avoid sensational images or claims that every person with these experiences should take the same action. Make it easy to find a helpline or emergency option without requiring disclosure in a public comment section. Where practical, remind readers to consider device privacy if monitoring is suspected. A publication cannot replace professional safety planning, but it can help someone recognise that a situation deserves attention and that support exists.

What is the difference between a survey victim and a police case?

A person may report abuse in a confidential survey without ever contacting police. Conversely, police may record several offences in one situation. ONS survey prevalence estimates and police-recorded offences therefore cannot be converted directly into one another. The survey concerns people; police figures concern recorded events under particular rules. Presenting 49,557 controlling or coercive behaviour offences as the number of all victims would be inaccurate. For the same reason, a decline in recorded crimes need not mean underlying experiences have declined at the same rate. This distinction is essential for responsible public reporting.

How might control develop gradually?

A partner may begin by requesting frequent check-ins, later insisting on seeing messages and eventually restricting contact with friends. Each step can be framed as love or protection. The issue is not that every check-in is abusive, but that the wider pattern may restrict independent choice and create fear of consequences. A person concerned about such changes can discuss them confidentially with a specialist service. An online list of red flags can raise awareness but should not be used as a substitute for a personal safety assessment, especially when the person is at risk of retaliation.

Can a financially dependent partner still have autonomy?

Yes. Unequal earnings, disability or caregiving arrangements do not automatically mean economic abuse. What matters is whether the individual can participate in decisions, access necessities and make choices without intimidation. Financial abuse may involve forced debt, restricted access or punishment for spending. These examples should be explored in context. A person who is economically dependent may face additional barriers to leaving, and public advice should recognise those constraints. Specialist support can help identify safe practical options rather than instructing someone to make sudden financial changes that increase danger.

How should someone support a disclosure?

Listen calmly, believe that the concern deserves attention and avoid interrogating the person to establish a perfect account. Ask what support would be useful and whether there are immediate safety risks. Do not promise absolute secrecy if someone faces imminent danger, but respect privacy and do not share details casually. A friend who demands immediate action may unintentionally increase isolation. It is more helpful to remain available, offer access to specialist services and understand that the person experiencing abuse may need to make choices gradually and under difficult constraints.

What should editorial opinion not do?

An editor may reasonably argue that control and intimidation deserve stronger prevention and clearer services, but opinion must not replace facts or make unsupported allegations about individuals. The ONS data support statements about population prevalence and recorded offences within defined periods, not conclusions about a reader’s specific partner. Strong editorial writing separates evidence, interpretation and practical guidance. Where personal safety is involved, uncertainty should be handled responsibly and directions should favour qualified support rather than sensational certainty. Precision is part of taking the issue seriously.

Support-seeking does not require a perfect label

A person may be unsure whether what they are experiencing meets a legal definition of coercive control. That uncertainty should not prevent them from asking a specialist organisation for confidential advice. Support workers can discuss patterns, risk and practical options without requiring someone to prove a criminal offence first. Friends should avoid turning conversations into cross-examinations. The priority is safety and increased independence. Population statistics show the scale of domestic abuse, but the needs of a particular individual should be assessed on the facts of their situation rather than on whether their experience resembles a common public example.

Evidence at a glance

ONS Crime Survey year ending March 2025: 3.8 million people or 7.8% experienced domestic abuse; emotional abuse 5.0%, economic abuse 2.7%, threats 2.6%. These may overlap. Police recorded 49,557 controlling/coercive behaviour offences, a different measure.

Illustrative real-world decision

Illustrative composite, not a real case: A person stops meeting friends because every outing causes angry accusations; later their partner demands access to banking and messages. They begin consulting a specialist helpline using a safe device and develop an individual plan rather than confronting the partner on the basis of a general internet article.

Editorial interpretation

Editorial assessment: Accurate prevalence data can establish the scale of a problem, but cannot replace careful attention to an individual's safety. A useful feature distinguishes survey estimates, police counts and real-world patterns of coercion while avoiding simplistic instructions that could increase risk.

Questions to ask and practical next steps

• What the latest comparable survey measures
• Coercive control is a pattern, not one argument
• The numbers on different types of abuse
• Economic abuse and daily dependence
• Technology-facilitated monitoring
• Sexual consent within an established relationship
• What police data can and cannot establish
• Recognising isolation and intimidation
• Finding support and deciding next steps